#microbenchmarks 検索結果

Well, #microbenchmarks are hard. ⏱️ Especially taking into account #GC-based environments. Here you can read about such experiment and why it was tricky to measure: bit.ly/2KFmiYz #elixirlang #elixir #myelixirstatus #benchee #garbagecollection #benchmark #benchmarks

afronski's tweet image. Well, #microbenchmarks are hard. ⏱️ Especially taking into account #GC-based environments. Here you can read about such experiment and why it was tricky to measure: bit.ly/2KFmiYz #elixirlang #elixir #myelixirstatus #benchee #garbagecollection #benchmark #benchmarks

„Microbenchmarks sind mittlerweile offizieller Bestandteil von OpenJDK“ bit.ly/2MpkpQm #Microbenchmarks #Java #Interview #bestpractices #jax2019 @Mr__Steel

JAXenter's tweet image. „Microbenchmarks sind mittlerweile offizieller Bestandteil von OpenJDK“ bit.ly/2MpkpQm #Microbenchmarks #Java #Interview #bestpractices #jax2019 @Mr__Steel

@TheMue and Python can be really slooow #microbenchmarks ;)


#Benchmarking is a #hairyscience in general, #microbenchmarks such as comparisons between versions of #raspberrypios (aka #raspbian) are an especially hairy situation as even the smallest influences can cause problems. I checked my stash of… 📸 instagram.com/p/CWogDyGNMP-/

tamhanna's tweet image. #Benchmarking is a #hairyscience in general, #microbenchmarks such as comparisons between versions of #raspberrypios (aka #raspbian) are an especially hairy situation as even the smallest influences can cause problems. I checked my stash of…

📸 instagram.com/p/CWogDyGNMP-/

Wer nicht weiß, was er misst, misst Mist! Gestern meistgelesen: Performanceoptimierung mit Microbenchmarks bit.ly/2Mc1fgR #Microbenchmarks #Performance #bestpractices

JAXenter's tweet image. Wer nicht weiß, was er misst, misst Mist! Gestern meistgelesen: Performanceoptimierung mit Microbenchmarks bit.ly/2Mc1fgR #Microbenchmarks #Performance #bestpractices

Finally, my @MeetupDE talk for the @jugbb will start in 2,5h at @LucaNet in Berlin. (in German) Already excited and a bit nervous - Topic is #Microbenchmarks #Java #JMH. If you cannot attend you'll also find most of the contents in the current issue of the @JavaMagazin (06/2019)


iPhone 4 CPU running @ 800MHz: http://bit.ly/5pxjgk 1.33x faster than 3GS. iPad 1.25x faster #iPhone #microbenchmarks


"I can build you a fast demo. The real question is 'Can I build you a fast app?'" - @richcampbell #microbenchmarks #dotnetrocks


Introducing #benchmark - a light and flexible C++ library for running #microbenchmarks @dominichamon buff.ly/LF0oBz


Array-backed Iterator is 7x faster than standard JDK8 ArrayList Iterator. #JMH #microbenchmarks #JVM #uberfastdata


📊 Benchmarks: MemoryStreamSlim outperforms both `MemoryStream` and RecyclableMemoryStream in real-world usage. See performance breakdowns ➡️ kzdev-net.github.io/kzdev.perfutil… #dotnet #perfmatters #microbenchmarks


C++ On Sea 2023 SESSION ANNOUNCEMENT: Employing modern C++ for high performance delta-coding compression by @thecppzoo cpponsea.uk/2023/sessions/… Register now at cpponsea.uk/tickets/ #microbenchmarks #reflection #catch2 #lowlatency #cpp


Our demo paper "#Microbenchmarks for Question Answering Systems Using QaldGen" has been accepted to #iswc_conf 2019. Demo available qaldgen.aksw.org. @DiceResearch @akswgroup @OyeKuldeep


"Are your optimizations making any improvement? A simple setup to benchmark two branches" cmdcolin.github.io/posts/2025-11-…


I benchmarked myself instead of my PC and discovered a lot bit.ly/4orDTRo

MUO_official's tweet image. I benchmarked myself instead of my PC and discovered a lot
bit.ly/4orDTRo

Minimax-M2 uses a 10B activated parameter design inside a 230B model, which keeps compute light while delivering good performance on long toolchains like mcp, shell, browser, retrieval, and code. It fits smoothly into workflows using Cursor, Cline, Kilo Code, Droid, and similar…

GithubProjects's tweet image. Minimax-M2 uses a 10B activated parameter design inside a 230B model, which keeps compute light while delivering good performance on long toolchains like mcp, shell, browser, retrieval, and code.
It fits smoothly into workflows using Cursor, Cline, Kilo Code, Droid, and similar…

Agreed, but I mostly mean that users should look into newer similarly sized models, which are all positioned to outperform all-MiniLM-L6-v2 with the same efficiency/speed/latency/cost: - huggingface.co/Snowflake/snow… - huggingface.co/ibm-granite/gr… - huggingface.co/mixedbread-ai/…


Benchmarked at 9 microseconds per trade 🤪 It's gonna become slower when I have to make the calls to the exchange, networking and all that shizzle... But I'm fooking proud of it! 😎


It won't be a single benchmark. Should look at aggregation like @ArtificialAnlys For coding swe bench pro and livebenchcode, for agentic perf METR is well respected


Done! I'm planning to run some comparisons too! I like benchmarks, but I trust them only if I can run them personally 😂

ivanfioravanti's tweet image. Done! I'm planning to run some comparisons too! I like benchmarks, but I trust them only if I can run them personally 😂

gMiden, Web3 family. Wednesday starts off more actively than Monday or Tuesday. The rocking has ended, and it's time for topics of an engineering nature. For example, the bench harness at @0xMiden. Miden bench harness is a shared performance benchmark. Here, builders test their…

MakassjakaMaks's tweet image. gMiden, Web3 family.
Wednesday starts off more actively than Monday or Tuesday.

The rocking has ended, and it's time for topics of an engineering nature. For example, the bench harness at @0xMiden.

Miden bench harness is a shared performance benchmark. Here, builders test their…

Tuesday is coming to an end, work is done, and we can now talk about something interesting. For example, how @0xMiden prepares for situations when something in the protocol may not go according to plan. Miden has a mechanism called guardian drills. These are predefined actions…

MakassjakaMaks's tweet image. Tuesday is coming to an end, work is done, and we can now talk about something interesting.

For example, how @0xMiden prepares for situations when something in the protocol may not go according to plan.

Miden has a mechanism called guardian drills.
These are predefined actions…


you can even run these benchmarks on your own its open source

asishcodes's tweet image. you can even run these benchmarks on your own 

its open source

btw 3% is actually a really big difference according to this benchmark

asishcodes's tweet image. btw 3% is actually a really big difference according to this benchmark


This was one hell of an undertaking. We still have a lot to learn with Linux benchmarking, but we finally have a foundation for it. This is our first-ever round of desktop Linux GPU gaming benchmarks -- we've previously only done this on handhelds: youtube.com/watch?v=ovOx4_…

GamersNexus's tweet card. RIP Windows: Linux GPU Gaming Benchmarks on Bazzite

youtube.com

YouTube

RIP Windows: Linux GPU Gaming Benchmarks on Bazzite


You can trust new benchmarks that haven’t been benchmaxxed yet: x.com/corby_rosset/s…

We are also releasing WebTailBench — a new benchmark covering underrepresented real-world tasks like booking tickets, comparing prices, and applying for jobs. Fara-7B scores 38.4% vs 25.7% for Operator and 19.5% for UI-TARS-1.5-7B.

corby_rosset's tweet image. We are also releasing WebTailBench — a new benchmark covering underrepresented real-world tasks like booking tickets, comparing prices, and applying for jobs. Fara-7B scores 38.4% vs 25.7% for Operator and 19.5% for UI-TARS-1.5-7B.


🦠 Antibiotic resistance = bacteria adapting to survive treatment — and spreading that resistance. MicroGenDX detects 17 resistance genes + 60K pathogens so providers can target the right bug with the right drug. 🧬 loom.ly/FmnRFhk #AntibioticResistance #Healthcare

microgendx's tweet image. 🦠 Antibiotic resistance = bacteria adapting to survive treatment — and spreading that resistance.

MicroGenDX detects 17 resistance genes + 60K pathogens so providers can target the right bug with the right drug.

🧬 loom.ly/FmnRFhk

#AntibioticResistance #Healthcare

🚀 New #LightBench release is out — our GPU benchmark for quick sanity checks and comparisons, built on the Light Tracer Render engine (now works on macOS too!)

hb3p8's tweet image. 🚀 New #LightBench release is out — our GPU benchmark for quick sanity checks and comparisons, built on the Light Tracer Render engine (now works on macOS too!)

I think this is why independent benchmarks like skatebench and snitchbench by @theo can prove to be better markers to understand model behaviors and people should learn how to test and educate their audiences.

I played with Gemini 3 yesterday via early access. Few thoughts - First I usually urge caution with public benchmarks because imo they can be quite possible to game. It comes down to discipline and self-restraint of the team (who is meanwhile strongly incentivized otherwise) to…



MiniMax-M2 3bit DWQ vs standard 3bit quant: +4.1 points on MMLU-Pro 📈 (Took a week to benchmark 😅) 🔗 huggingface.co/catalystsec/Mi…

kernelpool's tweet image. MiniMax-M2 3bit DWQ vs standard 3bit quant: +4.1 points on MMLU-Pro 📈
(Took a week to benchmark 😅)
🔗 huggingface.co/catalystsec/Mi…

openbenchmarking.org ここのCPUベンチ比較では、グラフの上に ms Fewer is Btterと表記されてるベンチ結果ではグラフの短い方が性能上で グラフの上にFPS More is Betterと表記されてるベンチ結果では、グラフの長い方が性能上


If we are being honest....this is the Benchmark!!!

TisBits's tweet image. If we are being honest....this is the Benchmark!!!
TisBits's tweet image. If we are being honest....this is the Benchmark!!!
TisBits's tweet image. If we are being honest....this is the Benchmark!!!
TisBits's tweet image. If we are being honest....this is the Benchmark!!!

Ain’t no fucking way SSR is gonna beat his own interval bang like RRR



I worked few weeks on tinybench. It is now about 3 kb smaller. also fully tested in @deno_land and @bunjavascript. + now without any external deps. Also performance of processing the samples should be significant faster and significantly less memory usage.


"#microbenchmarks" に一致する結果はありません

„Microbenchmarks sind mittlerweile offizieller Bestandteil von OpenJDK“ bit.ly/2MpkpQm #Microbenchmarks #Java #Interview #bestpractices #jax2019 @Mr__Steel

JAXenter's tweet image. „Microbenchmarks sind mittlerweile offizieller Bestandteil von OpenJDK“ bit.ly/2MpkpQm #Microbenchmarks #Java #Interview #bestpractices #jax2019 @Mr__Steel

Well, #microbenchmarks are hard. ⏱️ Especially taking into account #GC-based environments. Here you can read about such experiment and why it was tricky to measure: bit.ly/2KFmiYz #elixirlang #elixir #myelixirstatus #benchee #garbagecollection #benchmark #benchmarks

afronski's tweet image. Well, #microbenchmarks are hard. ⏱️ Especially taking into account #GC-based environments. Here you can read about such experiment and why it was tricky to measure: bit.ly/2KFmiYz #elixirlang #elixir #myelixirstatus #benchee #garbagecollection #benchmark #benchmarks

Wer nicht weiß, was er misst, misst Mist! Gestern meistgelesen: Performanceoptimierung mit Microbenchmarks bit.ly/2Mc1fgR #Microbenchmarks #Performance #bestpractices

JAXenter's tweet image. Wer nicht weiß, was er misst, misst Mist! Gestern meistgelesen: Performanceoptimierung mit Microbenchmarks bit.ly/2Mc1fgR #Microbenchmarks #Performance #bestpractices

#Benchmarking is a #hairyscience in general, #microbenchmarks such as comparisons between versions of #raspberrypios (aka #raspbian) are an especially hairy situation as even the smallest influences can cause problems. I checked my stash of… 📸 instagram.com/p/CWogDyGNMP-/

tamhanna's tweet image. #Benchmarking is a #hairyscience in general, #microbenchmarks such as comparisons between versions of #raspberrypios (aka #raspbian) are an especially hairy situation as even the smallest influences can cause problems. I checked my stash of…

📸 instagram.com/p/CWogDyGNMP-/

Loading...

Something went wrong.


Something went wrong.


United States Trends